Homo sapiens smoking data (GSE68559)
Run parameters
| Test | Expressioncutoff | Normalisation | Variancecutoff | Coexpression_measure | Edge_threshold | Threshold_method |
|---|---|---|---|---|---|---|
| GSE68559 | 1 | upper_quartile | 0.025 | pidc | 0.95 | empiricaldistzero |
Data preprocessing
Data cleaning
Before cut:
After cut:
Network construction
#manually copied from "/home/osbornejr/app/output/cache/GSE68559/expression_cutoff/20/normalisation/upper_quartile/sampling/0.025/similarity/pidc/sqrt_n/threshold/0.95/threshold_method/empirical_dist_zero"
# and output/cache/GSE68559/expression_cutoff/20/normalisation/upper_quartile/sampling/0.025/
using WGLMakie,JSServe,Markdown,Graphs,GraphMakie,NetworkLayout,JLD2,DataFrames
Page(exportable=true,offline=true)
WGLMakie.activate!()
sample_file = "_assets/menu1/sample_counts.jld2"
adj_file = "_assets/menu1/adjacency_matrix.jld2"
sample_counts = JLD2.jldopen(sample_file) do file
file["sample counts"]
end
pre_adj_matrix = JLD2.jldopen(adj_file) do file
file["pre-adj_matrix"]
end
adj_matrix = JLD2.jldopen(adj_file) do file
file["adjacency_matrix"]
end
components = JLD2.jldopen(adj_file) do file
file["components"]
end
#Trim nodes with degree zero
network_counts = sample_counts[components[1],:][vec(sum(pre_adj_matrix,dims=2).!=0),:]
#maintain list of vertices in graph
vertexlist = copy(network_counts[!,:transcript_type])
g = Graph(adj_matrix)
vertex_colors = replace(vertexlist,"noncoding"=>:red,"coding"=>:blue);
begin
set_theme!(backgroundcolor=:white)
fig,scene,p = graphplot(g;
layout=Spring(dim=3,C=1.0),
node_color = vertex_colors,
node_size = 5,
edge_color = :grey,
edge_width = 0.1,
figure = (resolution = (1500, 800),)
)
scene.show_axis =false
fig
end